well this is cool because me and my fellow neoliberal bros who love land value taxes are going to Addis for the jazz festival next year. good piece. i think most of the worries in it come apart on one distinction that's technically day-one micro and still gets underused all the time, even by economists who'd ace it on the exam: deadweight loss lives in marginal rates and substitution effects alone, while average rates, effective burdens, and lump-sum transfers move money without moving anyone's decision at the margin, so they cost nothing in efficiency. hold that firmly and most of these worries dissolve, because they're pricing transfers as if they were distortions.
so here's the mechanism I'd actually push on you, because it dissolves the valuation problem the lease exists to work around. harberger taxation: you self-assess, pay the rate on your own stated price, and stand ready to sell to anyone who meets it. that endogenizes the assessment, so there's no appraiser to game and no benchmark price to go stale and undershoot true value, which is the exact leak you describe as capturing only a fraction. the standard objection is that a forced sale can cost you the value your home has to you above what a buyer would pay. but freehold already does exactly this the moment you try to realize land appreciation, because the only way to cash it out is to sell, and no buyer loves your house the way you do. the fit between the place and your life is worth more to you than to the market, which is half the reason you're still living there. so you already surrender that surplus to move, and harberger only puts a visible price on a cost freehold was charging you in silence. it's a transfer freehold already imposed, and transfers don't distort. and because a tax's distortion scales with the produced share of the asset, land, being unproduced, can be captured toward 100% with essentially nothing left to distort. https://clayshentrup.medium.com/the-convergence-of-harberger-taxation-and-land-value-capture-how-destructive-rights-transform-10a824ecd53c
the speculation you flag dissolves the same way. a holder who prices the best-use date correctly and waits is earning a normal risk-adjusted return for bearing time and risk, and the community-created increment gets captured by the rent no matter who holds, so there's no efficiency loss to fix. blocking resale is the move that actually destroys value, by stranding parcels with owners who can't build.
and I'd want the collusion actually documented before treating it as a real drag on capture, because it's easy to assert and hard to sustain. model the whole ring as one party holding its highest internal valuation. the clearing price is set by the second-highest valuation in the field, so the ring holds the price down only if it contains the top two valuations in the entire auction. let it hold just the highest and the next outsider bids it right back to where open competition would have put it. so a ring has to corner the top of the value distribution, not just round up a lot of warm bodies, and the highest outside bidder is precisely the one it most needs to pull in and is least likely to. which means the whole defense is just advertising the sale to a wide enough pool. get enough independent bidders in the room and cornering the top two is infeasible, and there isn't much else you need to do about it. a reserve mops up the thin-field case where too few show, and continuous self-assessment is a reserve that's always on, with no scheduled clearing moment to coordinate a sit-out around. the collusion you cite reads as an artifact of a thin, discretionary, one-shot allocation, and says nothing about capture itself.
on the spending side, where you sound less sure, the same distinction settles it, this time on the goal, not the incentive. when people say they want progressive taxes, what they care about is a progressive schedule of effective rates. when they call a sales tax regressive, they mean effective rates too, since a sales tax is a flat marginal rate on consumption and only bites regressively once you measure it against income. marginal brackets are just the instrument they habitually grab to bend the effective curve, and they slide between the two framings depending on the conclusion they want, without noticing. hold them to effective rates consistently and a flat marginal rate plus a UBI is more progressive than the graduated schedule they favor, because the grant drives the bottom effective rate negative and no bracket structure does that. net of the grant your average rate is the flat rate minus grant-over-income, rising with income and negative at the bottom. ghenis works the income-tax version out in full. https://www.ubicenter.org/us-flat-tax and it costs nothing in efficiency, because deadweight loss rides the marginal rate, which never moved. the progressivity everyone actually wants lives entirely on the effective side, which is free; the marginal side, where the distortion would be, stays flat. that also means affordability was never bounded by efficiency: a UBI is only a redistribution mechanism, orthogonal to how much you route through it, so whatever you already spend on the safety net funds an equivalent grant, the same dollars minus the apparatus. https://medium.com/basic-income/if-we-can-afford-our-current-welfare-system-we-can-afford-basic-income-9ae9b5f186af and the apparatus is where the real waste sits: overhead, unclaimed benefits, one-size in-kind baskets, and above all implicit marginal rates so steep they clear 100% at the cliff, where a pennsylvania single mother nets more at $29k than at $69k. that implicit marginal rate is the deadweight loss, while the whole debate fixates on the size of the transfer, which costs nothing.
so the ideal just pushes the distinction to the corner. fund everything from land rent and pigouvian bases, zero and negative deadweight loss respectively, set the income tax's marginal rate to zero, and hand out a flat UBI. not one distorting bracket anywhere in the marginal schedule, income or land, while the effective rate still climbs steeply with wealth because land ownership is concentrated and the grant is equal. all the progressivity, none of the distortion, which only sounds paradoxical if you were pricing the effective rate as though it cost efficiency.
flat-plus-UBI also avoids a second cost buried in the marginal schedule: a convex bracket schedule isn't neutral to the time-profile or household structure of income. two people with identical lifetime earnings pay different lifetime tax if one earns lumpily and the other smoothly, because the fat years hit high brackets and the lean years never refund it, so you're taxing volatility on top of income. the annual means test has the same defect, and joint brackets manufacture marriage penalties on top. a flat marginal rate is neutral across all of it, only the lifetime total matters.
anyway. since you're an economics guy I'd love to get you pilled on approval voting next.
Very interesting. Thank you.
Seems like a lot of work compared to approximating land prices and charging people ~10% of that a year.
Possibly, but it has worked well in Singapore: https://progressandpoverty.substack.com/p/singapore-economic-prosperity-through
you should charge 100% of it, not 10%.
well this is cool because me and my fellow neoliberal bros who love land value taxes are going to Addis for the jazz festival next year. good piece. i think most of the worries in it come apart on one distinction that's technically day-one micro and still gets underused all the time, even by economists who'd ace it on the exam: deadweight loss lives in marginal rates and substitution effects alone, while average rates, effective burdens, and lump-sum transfers move money without moving anyone's decision at the margin, so they cost nothing in efficiency. hold that firmly and most of these worries dissolve, because they're pricing transfers as if they were distortions.
so here's the mechanism I'd actually push on you, because it dissolves the valuation problem the lease exists to work around. harberger taxation: you self-assess, pay the rate on your own stated price, and stand ready to sell to anyone who meets it. that endogenizes the assessment, so there's no appraiser to game and no benchmark price to go stale and undershoot true value, which is the exact leak you describe as capturing only a fraction. the standard objection is that a forced sale can cost you the value your home has to you above what a buyer would pay. but freehold already does exactly this the moment you try to realize land appreciation, because the only way to cash it out is to sell, and no buyer loves your house the way you do. the fit between the place and your life is worth more to you than to the market, which is half the reason you're still living there. so you already surrender that surplus to move, and harberger only puts a visible price on a cost freehold was charging you in silence. it's a transfer freehold already imposed, and transfers don't distort. and because a tax's distortion scales with the produced share of the asset, land, being unproduced, can be captured toward 100% with essentially nothing left to distort. https://clayshentrup.medium.com/the-convergence-of-harberger-taxation-and-land-value-capture-how-destructive-rights-transform-10a824ecd53c
the speculation you flag dissolves the same way. a holder who prices the best-use date correctly and waits is earning a normal risk-adjusted return for bearing time and risk, and the community-created increment gets captured by the rent no matter who holds, so there's no efficiency loss to fix. blocking resale is the move that actually destroys value, by stranding parcels with owners who can't build.
and I'd want the collusion actually documented before treating it as a real drag on capture, because it's easy to assert and hard to sustain. model the whole ring as one party holding its highest internal valuation. the clearing price is set by the second-highest valuation in the field, so the ring holds the price down only if it contains the top two valuations in the entire auction. let it hold just the highest and the next outsider bids it right back to where open competition would have put it. so a ring has to corner the top of the value distribution, not just round up a lot of warm bodies, and the highest outside bidder is precisely the one it most needs to pull in and is least likely to. which means the whole defense is just advertising the sale to a wide enough pool. get enough independent bidders in the room and cornering the top two is infeasible, and there isn't much else you need to do about it. a reserve mops up the thin-field case where too few show, and continuous self-assessment is a reserve that's always on, with no scheduled clearing moment to coordinate a sit-out around. the collusion you cite reads as an artifact of a thin, discretionary, one-shot allocation, and says nothing about capture itself.
on the spending side, where you sound less sure, the same distinction settles it, this time on the goal, not the incentive. when people say they want progressive taxes, what they care about is a progressive schedule of effective rates. when they call a sales tax regressive, they mean effective rates too, since a sales tax is a flat marginal rate on consumption and only bites regressively once you measure it against income. marginal brackets are just the instrument they habitually grab to bend the effective curve, and they slide between the two framings depending on the conclusion they want, without noticing. hold them to effective rates consistently and a flat marginal rate plus a UBI is more progressive than the graduated schedule they favor, because the grant drives the bottom effective rate negative and no bracket structure does that. net of the grant your average rate is the flat rate minus grant-over-income, rising with income and negative at the bottom. ghenis works the income-tax version out in full. https://www.ubicenter.org/us-flat-tax and it costs nothing in efficiency, because deadweight loss rides the marginal rate, which never moved. the progressivity everyone actually wants lives entirely on the effective side, which is free; the marginal side, where the distortion would be, stays flat. that also means affordability was never bounded by efficiency: a UBI is only a redistribution mechanism, orthogonal to how much you route through it, so whatever you already spend on the safety net funds an equivalent grant, the same dollars minus the apparatus. https://medium.com/basic-income/if-we-can-afford-our-current-welfare-system-we-can-afford-basic-income-9ae9b5f186af and the apparatus is where the real waste sits: overhead, unclaimed benefits, one-size in-kind baskets, and above all implicit marginal rates so steep they clear 100% at the cliff, where a pennsylvania single mother nets more at $29k than at $69k. that implicit marginal rate is the deadweight loss, while the whole debate fixates on the size of the transfer, which costs nothing.
so the ideal just pushes the distinction to the corner. fund everything from land rent and pigouvian bases, zero and negative deadweight loss respectively, set the income tax's marginal rate to zero, and hand out a flat UBI. not one distorting bracket anywhere in the marginal schedule, income or land, while the effective rate still climbs steeply with wealth because land ownership is concentrated and the grant is equal. all the progressivity, none of the distortion, which only sounds paradoxical if you were pricing the effective rate as though it cost efficiency.
flat-plus-UBI also avoids a second cost buried in the marginal schedule: a convex bracket schedule isn't neutral to the time-profile or household structure of income. two people with identical lifetime earnings pay different lifetime tax if one earns lumpily and the other smoothly, because the fat years hit high brackets and the lean years never refund it, so you're taxing volatility on top of income. the annual means test has the same defect, and joint brackets manufacture marriage penalties on top. a flat marginal rate is neutral across all of it, only the lifetime total matters.
anyway. since you're an economics guy I'd love to get you pilled on approval voting next.
Good job Wudu. Nice insight.